I agree there's no good scientific evidence against it; but we wouldn't really expect there to be, for the kinds of harm (long term psychological) some people hypothesise it to cause; very difficult to study, and as far as I know, nobody ever has studied it.
Actually, though it's not definitive, there have been a few small studies on attachment of infant and young children who were and weren't sleep-trained. IIRC, the sleeptrained babies were just as attached (no evidence of harm done by STing) and the parents were happier and rated the children's behavior as improved. Of course, none of these were long-term.
I have this general review bookmarked (this stuff is relevant to my job):
http://www.sciencedirect.com/science/article/pii/S1087079209001154"Whether managing infant sleep problems by minimizing parental interactive behaviors at bedtime leads to insecure attachment is unknown. Studies measuring proxies of insecure attachment (i.e., child behavior through the Child Behavior Checklist and maternal report of infant attachment through the Flint Infant Security Scale) suggest otherwise.[16], [17] and 18 However, definitive studies incorporating validated measures of attachment (e.g., Strange Situation Test) and physiological measures (e.g., cortisol) are warranted. Unfortunately, a recent attempt to conduct such a study in South Australia failed, due in part to community members opposed to behavioral interventions for infant sleep undermining recruitment (personal communication Dr Gradisar, Flinders University, South Australia). A second study of infants who took part in an earlier randomized controlled trial of behavioral strategies is underway with outcomes at child age six of child sleep, behavior, and salivary cortisol and parent mental health.19 We await these results with great interest."
also of interest:
"Persistent sleep problems in infancy predict maternal depression at child age two years.23 Managing infant sleep problems has been shown to consistently reduce maternal report of depression symptoms in uncontrolled and controlled trials.[12] and [24] Thus there is a large overlap between infant sleep problems and maternal psychopathology and evidence to support both directions in a transactional model i.e., that maternal mood may affect infant sleep and that infant night waking and problems settling may lead to maternal depression and anxiety. However, most studies have used maternal report of child sleep and maternal mental health leading to concerns about report bias. Depressed mothers may view their child's sleep more negatively but this almost certainly reflects the impact of the infant's sleep on the mother's life and her perceived need for help with her infant's sleep. Maternal report of infant sleep is a reliable indicator of infant sleep habits, as validated against overnight infra-red video recordings and actigraphy[25] and [26] but whether this holds true for depressed mothers is unknown."
